# LLMs Quick Learning Guide - Play Right with AI Workshop ## Repository Overview This is an open-source workshop teaching AI-driven development and testing workflows using Playwright. The workshop guides learners to become "AI Conductors" orchestrating automated development cycles. ## Core Concept Transform developers from "code writers" to "AI orchestrators" who direct AI tools to: 1. Generate applications 2. Write tests 3. Execute tests 4. Analyze failures 5. Self-repair issues ## Technical Stack - **Frontend**: HTML/CSS/JavaScript (React ready but unused) - **Testing**: Playwright for E2E testing - **AI Integration**: Claude, Gemini, GPT APIs - **Deployment**: GitHub Pages - **Language**: Traditional Chinese (繁體中文) with bilingual prompting ## Project Structure ``` /workshop/ - 8 complete chapters with exercises /prompts/ - 40+ golden prompts (version-agnostic) /sample-app-source/ - 4 example applications (TODO, Shopping, Multi-page, Capstone) /tests/ - E2E test suites /docs/ - Static website (port 8080) /integrations/ - AI service integrations /memory-bank/ - Project context persistence ``` ## Key Features 1. **Bilingual Prompting Strategy**: "Think in English, Output in Chinese" 2. **Playwright MCP Integration**: AI controls browser through natural language 3. **Self-Cycling Workflow**: Automated test-analyze-fix loop 4. **Progressive Learning**: From basic CRUD to complex systems ## Workshop Chapters 1. **AI Conductor** - Mindset shift & environment setup 2. **First Movement** - AI generates applications 3. **Second Movement** - AI as test strategist 4. **Third Movement** - AI writes Playwright tests 5. **Fourth Movement** - AI analyzes test failures 6. **Final Movement** - AI completes self-repair 7. **Variations** - Complex scenarios 8. **Capstone** - End-to-end challenge ## Sample Applications 1. **TODO App** (★★☆☆☆) - Basic CRUD with local storage 2. **Shopping List** (★★★☆☆) - Categories, budget tracking 3. **Multi-Page App** (★★★★☆) - Routing, forms, responsive 4. **Capstone Starter** (★★★★★) - Complete project template ## Development Commands ```bash npm run workshop:start # Start workshop on port 8080 npm run test # Run Playwright tests npm run validate:env # Validate environment npm run build:site # Build static site ``` ## AI Prompt Patterns ### Application Generation ``` You are a senior developer. Create a [type] application with: - [Feature 1] - [Feature 2] 請用繁體中文註解所有功能 ``` ### Test Strategy ``` 分析這個應用程式並設計測試策略: [Application code] Please focus on: - Critical user paths - Edge cases ``` ### Debug Analysis ``` Debug this test failure: [Error message] 請分析可能原因並提供修復建議 ``` ## Current Status - ✅ 8/8 chapters complete - ✅ 4/4 sample apps implemented - ✅ 40+ prompts (GPT version-agnostic) - ✅ E2E tests complete - ✅ GitHub Pages deployed - 🎯 UX/UI enhancement plan created (see docs/next-step.md) ## Next Steps (UX/UI Enhancement) **Phase 1 (Week 1-2)**: Glassmorphism design, navigation, mobile-first **Phase 2 (Week 3-4)**: Progress tracking, interactive playground, quizzes **Phase 3 (Week 5-6)**: Animations, gamification, performance optimization ## Key Files for LLMs 1. `/memory-bank/activeContext.md` - Current project state 2. `/docs/next-step.md` - UX/UI enhancement roadmap 3. `/PROJECT_STRUCTURE.md` - Complete file organization 4. `/CLAUDE.md` - AI assistant guidelines 5. `/.cursorrules` - Memory bank system ## Important Patterns - **TDD First**: All examples follow Test-Driven Development - **Incremental Complexity**: Each chapter builds on previous - **Practical Focus**: Real-world applications, not toy examples - **Error Learning**: Intentional bugs for debugging practice ## Success Metrics - User engagement: 70% chapter completion target - Performance: <3s load time on 3G - Accessibility: WCAG 2.1 AA compliant - Mobile usage: >50% of traffic ## Repository URLs - GitHub: https://github.com/clarencechien/play-right-with-ai - Live Demo: https://clarencechien.github.io/play-right-with-ai/ - Local Dev: http://localhost:8080 ## For AI Assistants When helping with this repository: 1. Maintain Traditional Chinese for all user-facing content 2. Use bilingual prompting (English technical, Chinese output) 3. Follow TDD principles for all code examples 4. Preserve the educational progression structure 5. Keep prompts version-agnostic (no GPT-4 specifics) ## Quick Start for Development ```bash git clone https://github.com/clarencechien/play-right-with-ai.git cd play-right-with-ai npm install npm run workshop:start # Runs on port 8080 ``` --- Last Updated: 2025-09-11 Workshop Version: 1.0.0 Status: Production Ready